Following nearly 10 yr of extensive laboratory evaluation, a vaccinia-rabies glycoprotein (V-RG) vaccine was the first recombinant virus to undergo limited North American field release on 20 August 1990. The free-ranging raccoon population on Parramore Island (Virginia, USA) was exposed to a high density (10 baits/ha) of vaccine-laden baits distributed on a 300 ha vaccination area. An annual total of 887 raccoons were live-trapped for sedation, physical examination and blood collection for rabies antibody determination; there was no evidence of adverse effects or lesions due to the vaccine. Age and sex distributions, mean body weights, and live-capture histories of raccoons from the vaccination and non-baited control areas were compared. There were no statistically significant differences in survivorship between the baited and non-baited areas, nor between rabies antibody-positive and antibody-negative raccoons from the vaccination area. There was no trend in field mortality that suggested an association with either tetracycline or sulfadimethoxine, used as biomakers, or with vaccine contact determined by antibody status. No gross or histopathologic lesions due to the vaccine were demonstrated among a subsample of live-trapped raccoons collected for gross necropsy, biomarker analysis, histopathologic examination, and V-RG virus isolation attempts. Recovery of V-RG virus was limited to the tonsils of two biomarker-positive, clinically healthy raccoons collected from the vaccination area for postmortem examination on days 2 and 4 following bait distribution. These data reinforce the extensive body of safety data on the V-RG virus and extend it to include field evaluation where vaccine is offered free-choice in abundance, in baits designed to attract free-ranging raccoons, in a relatively simple ecosystem.